Hotspots

Code ships faster than anyone can read it.

Know which functions to fix before your next incident. One command, any codebase — human-written or AI-generated.

TypeScript · JavaScript · Rust · Go · Python · Java · C · C# · Vue 

$ hotspots analyze scripts/

CRITICAL (20)
  17.91  enrich_issues.py:55   enrich_issues
  16.48  post_gate.py:110      lint
  14.80  ci.py:129            _run_ci
Install free View on GitHub

Install in 30 seconds

One command. No daemon. Works with defaults; tune thresholds, weights and excludes in .hotspotsrc.json when you need to.

$ brew install Stephen-Collins-tech/tap/hotspots
Run 
$ hotspots analyze scripts/
Output 
CRITICAL (20)
  17.91  enrich_issues.py:55   enrich_issues
  16.48  post_gate.py:110      lint
  14.80  ci.py:129            _run_ci

What you get — free

In under a minute, on any repo.

MIT licensed.

PR

Size up an AI-generated PR before it merges

hotspots diff reports changed lines inside touched functions — the strongest tested predictor of which PRs need a defect fix — alongside a structural risk view of every changed function.

CI

A hard stop on complexity spikes

Run in delta mode with --policy. CI fails if a PR — human or agent-authored — introduces a new critical function or spikes an existing one. No LLM makes that call; the threshold is deterministic math.

Score

Prioritize with a risk score

Every function gets a numeric score. No guessing — a ranked list you can act on in the next sprint.

Bugs

Find what's causing incidents

Combines cyclomatic complexity with churn history to surface functions that are hard to understand and frequently changing.

Team

Check coupling before you start

hotspots coordinate flags coupling and ownership risk for the files you're about to touch, from a file list, a diff, or staged changes.

Export

Triage in a spreadsheet

Export CSV or XLSX, or use --axes for separate Risk, Coupling and Ownership rankings.

Recent analyses

Browse all →

minio/minio

MinIO's cmd/ control paths carry the highest activity risk — 5 functions to address first

Oct 26, 2026

keras-team/keras

Keras's summary, export and RNN paths carry the highest activity risk

Oct 25, 2026

code-review-is-the-bottleneck

Code Is Free Now. Review Is the Bottleneck.

Oct 25, 2026

Highest risk

Full leaderboard →

mudler/LocalAI

LocalAI's static asset layer carries the highest activity risk

42.3

DioxusLabs/dioxus

DioxusLabs/dioxus: harness.rs carries the highest risk — 5 functions to address

41.1

redis/redis

redis/redis's module and networking layer carries the highest structural debt — 5 functions to address first

38.6

Real output. Real repos.

Every analysis on this site was run automatically against a real open-source repo — same tool, same command, nightly. You can reproduce any post locally in one command.

300+

repos analyzed

nightly

new analyses added

1 cmd

to reproduce any post

Reproduce any analysis locally 
$ hotspots analyze <path> --mode snapshot --format json --explain-patterns --force
Browse all analyses → Read the docs →

What should this become?

A newsletter, more API access, something else entirely — tell me what you'd actually use. No account, no required email, 30 seconds.

Tell me what you want →